GuacaMol
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| de novo molecular design | GuacaMol goal-directed tasks | Osimertinib MPO Score1 | 23 | |
| Molecule generation | GuacaMol | Validity100 | 20 | |
| Unconditional molecule generation | GuacaMol | Validity100 | 20 | |
| Molecule Graph Generation | GuacaMol (test) | Validity100 | 17 | |
| Multi-property conditional molecule generation | GuacaMol | Unique Molecules430.8 | 16 | |
| Molecular Generation | GuacaMol (test) | Amlo (AUC-Top10)0.781 | 10 | |
| Molecule Optimization | GuacaMol v1 | Med1 Score39.9 | 8 | |
| Molecule Optimization | GuacaMol (test) | Total Score Sum4.903 | 8 | |
| Distribution-learning | GuacaMol (test) | Validity100 | 8 | |
| Time-varying molecular optimization | GuacaMol (ell_w = 0.3) | Average Rank1.9 | 7 | |
| Time-varying molecular optimization | GuacaMol ell_w = 0.2 | Average Rank1.75 | 7 | |
| Time-varying molecular optimization | GuacaMol ell_w = 0.1 | Average Rank1.98 | 7 | |
| Molecular Optimization | GuacaMol | Median Score 10.242 | 7 | |
| Molecular Graph Generation | GuacaMol | Validity100 | 6 | |
| Molecular Generation | GuacaMol QED top-100 | Mean Score0.948 | 5 | |
| Topological Preservation Analysis | GuacaMol | Kappa FR0.862 | 2 | |
| Conditional Molecule Generation | GuacaMol | Validity98.1 | 2 |
Showing 17 of 17 rows